Building an affordable housing application starts with data
Missed our previous posts? Here’s an introductory article about our Code for America fellowship. Rohan’s earlier post explaining Austin’s affordable housing situation with some data insights can be found here.
Our involvement with this affordable housing project started with a housing search app. Rohan heard about Austin’s affordable housing issue at a city council meeting, and soon after he chose it as his subject matter at a Data for Democracy hackathon. There, Rohan used publicly available affordable housing data provided by the City of Austin to build out a website that displayed unit information on a map of the city. He brought me in on the project for some help with the design and UX of the application.
At the end of the hackathon, we shared the affordable housing site with the government agencies and partners in the affordable housing space. Everyone was excited about the potential of the site, but we soon discovered one fundamental issue that made the site unusable in its current state — the reliability of the data.
Turns out, the affordable housing data made publicly available by the city is incomplete and out of date. Within Austin city limits, there are two government agencies and multiple non-profit partners that work on affordable housing. Each of these organizations has their own data set that is housed in a spreadsheet and not regularly maintained. The data we used for the search application was from just one of these organizations and hadn’t been updated in months.
This discovery led to a pivot in the trajectory of our project. Even though we ultimately want to help the people searching for affordable housing, we know this can’t be done until we have a reliable source of data behind the scenes. From research and interviews with residents searching for affordable housing, we found that the most difficult part of the search process is finding out which units are actually available. Currently, this involves looking at a list of properties on a sheet of paper and calling each prospect to find out their availability. This can be frustrating, tedious, and time-consuming.
With this in mind, we split our fellowship into three phases, outlined in the following diagram.
The first phase of our fellowship was spent gathering data sets from each affordable housing organization, combining them, and deduplicating them. We now have one master data set that we can use going forward.
The second phase is to build a portal to access and update this master data set. We envision that this portal will be used by anyone who works with properties in the affordable housing space. Instead of maintaining separate spreadsheets of affordable housing data within each organization, there will be one shared place for this data to live. Mock ups of the data portal application are below.
Once we have a reliable set of data about affordable housing, we will provide access to it through an API. This opens the door for the data to be used by any applications or websites that can utilize it. The third and final phase of our fellowship will be to use the API to build out the resident-facing housing search application, shown below.